the native libs offer a cleanup method, but this is not jni'd out by the gvsig libs. anyway this would need to be triggered. maybe by a specific clean image layer cache plugin, but i doubt this is worth the effort.
when i have time i will still look into a proper cache cleanup for closing ecw sdk layers. ede On 03.10.2011 09:40, Rahkonen Jukka wrote: > Hi, > > The memory reserved by java really goes down after about half an hour idle > period. > > -Jukka- > > edgar.soldin wrote: > >> Yes, ecw code seems to access the files directly through the >> native code and new images are created for the view >> requested. I found this interesting bit though >> http://iws.erdas.com/forum/memory-leak.aspx >> >> jukka could you try to fill your memory up a bit and wait for >> say 30 minutes and see if the memory reduces if nothing is requested? >> >> The explicit file closing with true for cache cleanup on >> layer removal is something i think about putting in, though i >> am still not sure where to hook in. >> >> ..ede >> >> On 29.09.2011 18:54, Stefan Steiniger wrote: >>> this answer probably doesn't help as i do not know the gvSIG code, >>> and is more of a note: >>> >>> I think for mrsid code created images in a temporary sub >> folder of oj. >>> Processing in sextante also writes rasters to a folder. For >> intermediate >>> results these are deleted (if the code is written), but >> final processing >>> results are stored in windows temp (and the user needs to >> clean them >>> up). Though... ECW may work completely differently >>> >>> On 29/09/2011 7:24 AM, edgar.sol...@web.de wrote: >>>> i can reproduce this, actually the memory is not even >> freed when the layer is removed from the task. >>>> >>>> removing layers seems to work like this >>>> >>>> - RemoveLayerPlugin execute >>>> - fill layer with empty feature collection >>>> - remove layer from layermanagement >>>> >>>> problem is: there is no active cleanup, the best i found >> was a finalize() method in the ecw java binding code >>>> >>>> src-ecw/com/ermapper/ecw/JNCSFile.java >>>> protected void finalize() >>>> throws Throwable >>>> { >>>> if(bIsOpen) >>>> ECWClose(true); >>>> super.finalize(); >>>> } >>>> >>>> but it seems like oj never reaches there, or better i set >> a debug point there but gc'ing does not reach it. >>>> >>>> any ideas, anyone? >>>> >>>> >>>> ede >>>> >>>> PS: i fetched the jecw java/native sources from gvsig, so >> we are gpl compliant by having it available at least in our svn >>>> >>>> >>>> On 28.09.2011 16:18, Jukka Rahkonen wrote: >>>>> Hi, >>>>> >>>>> There is one 12000x12000 JPEG2000 image for you at >>>> SNIP >>>>> It is lossless and therefore still 181 MB in size. Open >> it with OJ, zoom >>>>> and pan wildly around and you will see how java.exe will >> take more and >>>>> more memory. >>>>> >>>>> -Jukka- >>>>> >>>> >> -------------------------------------------------------------- >> ---------------- >>>> All the data continuously generated in your IT >> infrastructure contains a >>>> definitive record of customers, application performance, security >>>> threats, fraudulent activity and more. Splunk takes this >> data and makes >>>> sense of it. Business sense. IT sense. Common sense. >>>> http://p.sf.net/sfu/splunk-d2dcopy1 >>>> _______________________________________________ >>>> Jump-pilot-devel mailing list >>>> Jump-pilot-devel@lists.sourceforge.net >>>> https://lists.sourceforge.net/lists/listinfo/jump-pilot-devel >>> >>> >>> >> -------------------------------------------------------------- >> ---------------- >>> All the data continuously generated in your IT >> infrastructure contains a >>> definitive record of customers, application performance, security >>> threats, fraudulent activity and more. Splunk takes this >> data and makes >>> sense of it. Business sense. IT sense. Common sense. >>> http://p.sf.net/sfu/splunk-d2dcopy1 >>> _______________________________________________ >>> Jump-pilot-devel mailing list >>> Jump-pilot-devel@lists.sourceforge.net >>> https://lists.sourceforge.net/lists/listinfo/jump-pilot-devel >> >> -------------------------------------------------------------- >> ---------------- >> All of the data generated in your IT infrastructure is >> seriously valuable. >> Why? It contains a definitive record of application >> performance, security >> threats, fraudulent activity, and more. Splunk takes this >> data and makes >> sense of it. IT sense. And common sense. >> http://p.sf.net/sfu/splunk-d2dcopy2 >> _______________________________________________ >> Jump-pilot-devel mailing list >> Jump-pilot-devel@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/jump-pilot-devel >> > ------------------------------------------------------------------------------ > All the data continuously generated in your IT infrastructure contains a > definitive record of customers, application performance, security > threats, fraudulent activity and more. Splunk takes this data and makes > sense of it. Business sense. IT sense. Common sense. > http://p.sf.net/sfu/splunk-d2dcopy1 > _______________________________________________ > Jump-pilot-devel mailing list > Jump-pilot-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/jump-pilot-devel ------------------------------------------------------------------------------ All the data continuously generated in your IT infrastructure contains a definitive record of customers, application performance, security threats, fraudulent activity and more. Splunk takes this data and makes sense of it. Business sense. IT sense. Common sense. http://p.sf.net/sfu/splunk-d2dcopy1 _______________________________________________ Jump-pilot-devel mailing list Jump-pilot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/jump-pilot-devel
